Indications, mortality, and long-term outcomes of 50 consecutive patients undergoing damage control laparotomy for abdominal gunshot wounds

Khaled Twier¹, Leila Hartford¹, Andrew Nicol¹, Sorin Edu¹, Derek Roberts², Chad Ball², Pradeep Navsaria^1*

¹Trauma Center, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa

²Trauma and Acute Care Surgery, University of Calgary, Calgary, Canada

Received: July 26, 2019       Revised: August 31, 2019         Accepted: September 23, 2019

Introduction

Firearm injuries represent the sixth leading mechanism of injury managed by the Cape Town trauma system, Cape Town, South Africa (1-8). The mortality rate associated with abdominal gunshot wounds (GSWs) remains high with many deaths occurring as a result of exsanguination and irreversible shock (9-12). Damage control laparotomy (DCL) has increasingly been utilized for the management of severely injured patients with abdominal injuries (13-15). The DCL is characterized by an abbreviated initial operation that aims to rapidly and effectively control hemorrhage and contamination (13, 16-18).  Moreover, DCL has been reported to be associated with improved survival at the expense of increased morbidity. (13, 17, 18).

Regarding several aspects of DCL, there exist considerable uncertainties. The modern-day outcomes of patients subjected to DCL for abdominal GSWs remains relatively unknown in spite of the improvements in understanding DCL complications, pre-hospital and surgical care, resuscitation strategies, adjunctive pharmacotherapy, and other aspects of injury care in recent years (12, 19, 20-26). There is currently little evidence regarding which variable may reliably predict morbidity and mortality in patients managed with damage control laparotomy for abdominal GSWs (27-31). This is especially true as many previous studies enrolled patients injured by a variety of mechanisms in addition to abdominal GSWs, several of which have been reported to be associated with different outcomes (27-31). This study aimed to evaluate the clinical characteristics as well as short- and long-term outcomes of patients managed with damage control laparotomy for abdominal GSWs.

Methods

This retrospective study was conducted on patients who underwent a damage control laparotomy for abdominal GSWs at the Trauma Center of Groote Schuur Hospital (GSH), Cape Town, South Africa, between August 1^st, 2004 and September 30^th, 2009.

No formal a priori sample size calculation was conducted in this study. The number of eligible patients during the study period determined the sample size. Patients who underwent damage control laparotomy for an abdominal GSW were identified and screened for eligibility. One of the authors reviewed the hospital records of eligible patients to extract the required data after obtaining ethical approval.

Variables of interest were collected using a standardized data extraction form in the Microsoft Excel spreadsheet. These variables include 1) age, 2) gender, 3) vital signs, 4) Glasgow Coma Scale Scores, 5) number and location of abdominal GSWs, 6) whether more than one anatomical region or cavity was injured, 7) the results of pre- and intraoperative diagnostic imaging and arterial blood gas and laboratory tests, 8) core body temperatures of patients in the trauma bay and operating theatre (OT) settings, 9) whether patients were reported to have a clinically-observed coagulopathy in the OT, 10) the volume or units of crystalloids, packed red blood cells (PRBCs), fresh frozen plasma, platelets, and cryoprecipitate administered across the trauma bay and OT settings, 11) whether tranexamic acid was given or a thoracotomy was required, 12) the time interval between patient arrival in the trauma bay and the initiation of the operative procedure, 13) the exact injury pattern identified, 14) the operative procedures conducted, 15) the number of abdominal operations performed, and 16) the length of time from laparotomy to the closure of the abdominal wound.

Where possible, the data utilized in the indications for damage control laparotomy were collected in this study. Moreover, Injury Severity Scale (ISS), Revised Trauma Score (RTS), Trauma Related Injury Severity Score (TRISS), and Penetrating Abdominal Trauma Index (PATI) scores were calculated for each patient.

Outcomes of interest included the risk of mortality in the OT and within 6- or 24-hours and 30-days of operation. Morbidity was evaluated by collecting data on complications, including the open abdomen, intra-abdominal abscesses/sepsis, and entero-atmospheric fistulae. The lengths of hospital and intensive care unit stay and the long-term risk of mortality, small bowel obstruction, and ventral hernia were also determined in this study.

Continuous variables were grouped to form new categorical variables, based on previous studies and clinical relevance. The clinical characteristics and outcomes of patients who underwent damage control laparotomy for abdominal GSWs will be summarized using proportions, mean±SD, and medians with interquartile ranges (IQRs) as appropriate. Initial data exploration was conducted using cross-tabulation and the Mantel-Haenszel test. Following this, logistic regression was used to compute the unadjusted estimates with 95% confidence intervals for the association between each of the variables and the odds of survival after DCL for abdominal GSWs. Variables for the inclusion in the multivariate logistic regression were selected based on previous literature and clinical grounds. These variables included proxies for the “Lethal Triad” that were found to be associated with the outcome on univariate analysis (i.e., hypothermia [temperature dichotomized at 32 degrees Celsius], acidosis [lactate dichotomized
at 8mmol/L intraoperative], and coagulopathy [massive transfusion of greater than 10 units PRBCs]).

Furthermore, the trauma scoring system was included with the strongest association between mortality and PATI. Age was considered as a priori confounder. The data were analyzed in Stata MP (version 13.1, Stata Corp., College Station, TX, U.S.A). Proportions, odds ratios, 95% confidence intervals, and p-values are reported to two decimal places. All p-values in the logistic regression models were derived from the Likelihood Ratio Test. 

The study protocol was approved by the Human Research Ethics Committee of the University of Cape Town, Cape Town, South Africa, in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards (UCT HREC ref.:455/2016).

Since this study was conducted based on a retrospective chart/database review design, no consent was taken from the patients, and they were grouped according to standard unit protocols. All data were collected by the first and last authors on a password-protected computer and Excel sheet. None of the patients could be identified in the database as they were numbered consecutively from 1-50. It should be noted that the participants were all informed of the privacy and confidentiality of data.

Results

In total, 82 potentially eligible patients were identified from the GSH trauma database prior to the submission of the study protocol for ethics review. Out of this population, 32 cases were excluded from the study due to the following reasons:

·   Patients' death in the emergency room prior to being taken to the operating room (n=12)

·   Lack of damage control laparotomy (n=10)

·   Damage control laparotomy for injuries other than GSWs to the abdomen (n=4)

·   Missed patient records (n=2)

·   Erroneously duplicated patient folder numbers (n=4)

Accordingly, this study included 50 patients (48 male and 2 female) who underwent DCL for abdominal GSWs. The mean age of the participants was obtained at 29.7±10.88 years (age range: 17-62 years). Totally, 30 (60%) and 20 (40%) patients had single and multiple gunshot wounds to the abdomen, respectively. The mean values of preoperative systolic blood pressure, heart rate, respiratory rate, temperature, and Glasgow Coma Scale were 90.4 mmHg, 112 beats/per min, 19 breaths/per min, 35.6 degrees Celsius, and 12 with an IQR of 13 to 15, respectively.

According to the results obtained from pre-operative blood investigations, the mean values of hemoglobin concentration, pH, base deficit, and lactate were estimated at 8.35g/dl, 7.21, -9.71 mEq/L, and 5.25 mmol/L, respectively. This study computed four trauma indices, including ISS, RTS, TRISS, and PATI with the mean values of 27.12, 6.65, 82.36%, and 34.5, respectively. Intra-operatively, the mean values of temperature, pH, base deficit, and lactate were obtained at 35 degrees Celsius, 7.07, -14.41 mEq/L, and 7.95 mmol/L, respectively (Table 1).

The index procedure was terminated due to the presence of metabolic acidosis (n=46), increasing ionotropic support (n=39), rising lactate (n=34), and diffuse hemorrhage (n=28) in the patients. Hypothermia was noted as a reason for the termination of the index procedure (n=16), and disseminated intravascular coagulopathy (n=16).

Most patients received either two (42% of the patients) or three (38% of the patients) liters of crystalloids and one (34% of the patients) to two (40% of the patients) units of packed red cells pre-operatively. In addition, most patients received intra-operatively between one and six litres of crystalloids (78% of the patients), four to ten units of packed red cells (60% of the patients), between two and six units of fresh frozen plasma (85% of the patients), and one unit of platelets (32% of the

patients) (Table 2). 

In total, 40 (80%) patients had extra-abdominal GSWs. The GSWs were located at  posterior torso (n=24, 48%), epigastrium (n=12, 24%), right iliac fossa (n=12, 24%), left iliac fossa (n=8, 16%), right upper quadrant (n=6, 12%), and left upper quadrant (n=5, 10%).

Out of 50 patients, five cases underwent thoracotomy in the operating room at the time of laparotomy. The most common injuries noted at laparotomy were liver (28%), small bowel (44%), major venous injuries (40%), colon (38%), major arterial injuries (22%), and pancreas (20%) (Table 3).

The mean time for operation and the mean duration of surgery were 141±80.3 and 141±59.8

min, respectively. In total 27 (54%) patients died in the first six hours (n=14),  between six and 24 hours (n=60),  between 24 hours and seven days (n=3), and between seven to 30 days (n=4).  Moreover, out of 27 deaths, 1,10, and 16 cases occurred in the operating room,  the ward after ICU discharge, and the ICU, respectively. The mean length of stay in the ICU amongst the patients who survived the index procedure was 7.64±7.90 days. Out of those patients who survived ICU, the mean stay was 12.38±10.22 days. Moreover, the mean total length of hospital stays amongst those patients who survived the index procedure was 16.64 ±15.33 days.

The most common short-term complications documented in the ICU were an acute renal failure (n=10), multi-organ dysfunction (n=10), DIC (n=7), and abdominal compartment syndrome (n=7). Other complications included prolonged intubation requiring tracheostomy, ventilator-associated pneumonia, superficial wound sepsis, intra-abdominal sepsis, anastomotic leak, deep vein thrombosis, limb ischemia, hypoglycemia, and gastrointestinal bleeding.

Amongst those patients who survived ICU, the most common complications noted in the ward were superficial wound sepsis (n=5), open abdomen (n=5), paralytic ileus (n=4), and wound dehiscence (n=4). Other complications included atelectasis, pleural effusion, acute kidney injury, pancreatic fistula, and a high output ileostomy.

Table 4 summarizes the Clavien-Dindo Classification of all complications mentioned above. Long-term complications were noted in seven patients, and three patients presented with adhesive small bowel obstruction at 3 months, 1 year, and seven years, respectively, after the index operation. Furthermore, three patients presented with elective ventral hernia repair at 9 months, 1 year, and 6 years, after the index operation. One patient known with a ventral hernia lost the follow-up.

Univariate analysis of the association between the variables is listed in Tables 5 and 6. The mortality rate revealed strong evidence that PATI of less than 25, pre-operative infusion of greater than 3 liters of crystalloids, intra-operative lactate level of less than 8mmol/L, and intra-operative transfusion of fewer than 10 units of PRBCs increased the odds of survival.

In this study, there is significant evidence (P=0.01) that the odds of survival are 81% lower in

patients with a PATI of more than 25, compared to those with a PATI of less than 25 (OR 0.19, 95% CI: 0.05-0.72).

There is also significant evidence (P=0.02) that patients who received more than three liters of crystalloids pre-operatively were four times more likely to survive, compared to those who received less than three liters of crystalloids (OR 4.12, 95% CI: 1.23-13.77). On univariate analysis, there is significant evidence (P=0.01) that patients with an intraoperative lactate level of greater than 8 were 82% less likely to survive (OR 0.18, 95% CI: 0.05-0.6), and those who had massive transfusion (≥10 units) of PRBCs intra-operatively were 78% less likely to survive (OR 0.22, 95% CI: 0.05-0.92). In addition, according to the multivariate analysis results, there is significant evidence (P=0.04) that after considering the confounding effect of age and all other variables in the regression model, a PATI of greater than 25 was associated with an 80% lower odds of survival after DCL (OR 0.20, CI 0.04-1.00); however, there is considerable uncertainty around this estimate owing to the small sample size.

Discussion

This study aimed to describe the clinical characteristics of patients managed with DCL for abdominal GSWs at GSH. The indications and long-term outcomes of patients undergoing DCL were also examined in this study.  The overall mortality rate after DCL was obtained at 54%. Univariate analysis revealed that PATI of less than 25, pre-operative infusion of greater than 2 liters of crystalloids, intra-operative lactate level of less than 8mmol/L, and intra-operative transfusion of fewer than 10 units of PRBCs associated with an increase in the odds of survival after DCL.

In the same line, the multivariate analysis revealed significant evidence (P=0.04) that a PATI score of greater than 25 was associated with an 80% decrease in the odds of survival (OR 0.20, 95% CI: 0.04-1.00). After considering the confounding effect of age and all other variables in the model, there was no evidence of a correlation between intra-operative lactate and temperature or massive transfusion with the odds of survival after DCL.

Given the substantial uncertainty around the estimates demonstrated by wide confidence intervals for most of the odds ratios in the univariate analysis and most likely the small sample size, the results summarized above should be interpreted with caution.

The multivariate regression controls the confounding effect of the other variables in the model; however, the observed association between a high PATI score and a decrease in the odds of survival should not be interpreted as a causal relationship since the results may still be subject to random sampling error, selection bias, information bias, or residual confounding after analysis.

The mortality rate after DCL for abdominal GSWs was notably higher in this study than in previously published studies conducted by Rotondo et al. (18) (42%) and Brennar et al. (32) (28%). Factors contributing to this could anecdotally include the delay to operation and long operation time, as well as other resource-based constraints, such as the level of training of healthcare workers, access to the operating room and intensive care, as well as the availability of allied healthcare support. Not captured in the data presented here is the additional delay between injury and arrival time at GSH owing to ambulance and emergency service shortages.

The results of the univariate analyses support the hypothesis that the “Lethal Triad” (i.e., coagulopathy, acidosis, and hypothermia) contributes significantly to patient mortality following major trauma. In this study, high levels of lactate and massive transfusion were associated with a negative outcome. Frischknecht et al. identified several risk factors for early mortality, including severe head injury and the “Lethal Triad” in patients undergoing damage control procedures. (30) Similarly, Timmermans et al. demonstrated that
a large base deficit, low pH, hypothermia, coagulopathy, and massive transfusion correlated with mortality in trauma patients. (27)

The incidental finding of an association between high volumes of crystalloids administered pre-operatively and an increased odds of survival should be interpreted with caution, given the wide confidence interval of the estimate probably because of the small sample size. Due to the lack of corroborative evidence, and given that the analysis is based on relatively small numbers, it is not recommended to implement any change in fluid prescribing practice on the strength of these results.

It is recommended that future studies examine the effect of the ratio of crystalloids to blood products on the odds of survival after DCL since it is hypothesized that initial permissive hypotension followed by definitive control of hemorrhage and blood product based resuscitation is associated with positive outcomes.

The multivariate analysis demonstrated significant evidence that high PATI scores are associated with increased mortality after controlling age and intraoperative factors. This finding is consistent with the results of a similar study performed by Adesanya et al. who demonstrated that a PATI score of >15 was associated with a 20-fold increase in the incidence of death (P<0.0001) (33).

There are many factors that may impact the external validity of the observed results. Notably, data on the socio-economic status, immune profile, and comorbid conditions were not collected or analyzed in this study. These factors may have a significant impact on patients’ recovery after major trauma. Therefore, the observed results may not be applicable to high-income countries with different patient demographics.

Conclusions

In this limited study, the overall mortality rate of patients that required DCL for abdominal GSWs was obtained at 54%.  There is significant evidence that after controlling confounding variables, the PATI score of >25 associated with decreased odds of survival (OR: 0.20, P=0.04).

Acknowledgments

There is nothing to declare.

Funding

There is no funding source.

Conflict of Interest

The authors declare that they have no conflict of interest regarding the publication of the study.

References

1.     Nicol A, Knowlton LM, Schuurman N, Matzopoulos R, Zargaran E, Cinnamon J, et al. Trauma surveillance in cape town, south africa: an analysis of 9236 consecutive trauma center admissions. JAMA Surg. 2014; 149(6):549-556. PMID: 24789507 DOI: 10.1001/jamasurg.2013.5267

2.     Navsaria PH, Edu S, Nicol AJ. Civilian extraperitoneal rectal gunshot wounds: surgical management made simpler. World J Surg. 2007; 31(6):1345-51. PMID: 17457641 DOI: 10.1007/s00268-007-9045-z

3.     Navsaria PH, Nicol AJ. Selective nonoperative management of kidney gunshot injuries. World J Surg. 2009; 33(3):553-7. PMID: 19132440 DOI: 10.1007/s00268-008-9888-y

4.     Navsaria PH, Nicol AJ, Krige JE, Edu S. Selective nonoperative management of liver gunshot injuries. Ann Surg. 2009; 249(4):653-6. PMID: 19300222 DOI: 10.1097/SLA.0b013e31819ed98d

5.     Chinnery GE, Krige JE, Kotze UK, Navsaria P, Nicol A. Surgical management and outcome of civilian gunshot injuries to the pancreas. Br J Surg. 2012; 99(Suppl 1):140-8. PMID: 22441869 DOI: 10.1002/
bjs.7761

6.     Vassar MJ, Kizer KW. Hospitalizations for firearm-related injuries. A population-based study of 9562 patients. JAMA. 1996; 275(22):1734-9. PMID: 8637170 DOI: 10.1016/S0140-6736(11)60512-6

7.     Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: global burden of disease study. Lancet. 1997; 349(9061):1269-76. PMID: 9142060 DOI: 10.1016/S0140-6736(96)07493-4

8.     Gore FM, Bloem PJ, Patton GC, Ferguson J, Joseph V, Coffey C, et al. Global burden of disease in young people aged 10-24 years: a systematic analysis. Lancet. 2011; 377(9783):2093-102. PMID: 21652063 DOI: 10.1016/S0140-6736(11)60512-6

9.     Nance FC, Wennar MH, Johnson LW, Ingram JC Jr, Cohn I Jr. Surgical judgment in the management of penetrating wounds of the abdomen: experience with 2212 patients. Ann Surg. 1974; 179(5):639-46. PMID: 4823841 DOI: 10.1097/00000658-197405000-00017

10.  Payne JE, Berne TV, Kaufman RL, Dubrowskij R. Outcome of treatment of 686 gunshot wounds of the trunk at Los Angeles County-USC Medical Center: implications for the community. J Trauma. 1993; 34(2):276-81. PMID: 8459470 DOI: 10.1097/
00005373-199302000-00018

11.  Feliciano DV, Burch JM, Spjut-Patrinely V, Mattox KL, Jordan GL Jr. Abdominal gunshot wounds. An urban trauma center's experience with 300 consecutive patients. Ann Surg. 1988; 208(3):362-70. PMID: 3421760 DOI: 10.1097/00000658-198809000-00014

12.  Duchesne JC1, McSwain NE Jr, Cotton BA, Hunt JP, Dellavolpe J, Lafaro K, et al. Damage control resuscitation: the new face of damage control. J Trauma. 2010; 69(4):976-90. PMID: 20938283 DOI: 10.1097/TA.0b013e3181f2abc9

13.  Stone HH, Strom PR, Mullins RJ. Management of the major coagulopathy with onset during laparotomy. Ann Surg. 1983; 197(5):532-5. PMID: 6847272 PMCID: PMC1353025 DOI: 10.1097/00000658-198305000-00005

14.  Hatch QM, Osterhout LM, Ashraf A, Podbielski J, Kozar RA, Wade CE, et al. Current use of damage-control laparotomy, closure rates, and predictors of early fascial closure at the first take-back. J Trauma. 2011; 70(6):1429-36. PMID: 21817981 DOI: 10.1097/TA.0b013e31821b245a

15.  Martin MJ, Hatch Q, Cotton B, Holcomb J. The use of temporary abdominal closure in low-risk trauma patients: helpful or harmful? J Trauma Acute Care Surg. 2012; 72(3):601-6. PMID: 22491542 DOI: 10.1097/TA.0b013e31824483b7

16.  Roberts DJ, Zygun DA, Kirkpatrick AW, Ball CG, Faris PD, Bobrovitz N, et al. A protocol for a scoping and qualitative study to identify and evaluate indications for damage control surgery and damage control interventions in civilian trauma patients. BMJ Open. 2014; 4(7):e005634. PMID: 25001397 DOI: 10.1136/
bmjopen-2014-005634

17.  Burch JM, Ortiz VB, Richardson RJ, Martin RR, Mattox KL, Jordan GL Jr. Abbreviated laparotomy and planned reoperation for critically injured patients. Ann Surg. 1992; 215(5):476-83. PMID: 1616384 DOI: 10.1097/00000658-199205000-00010

18.  Rotondo MF, Schwab CW, McGonigal MD, Phillips GR 3rd, Fruchterman TM, Kauder DR, et al. 'Damage control': an approach for improved survival in exsanguinating penetrating abdominal injury. J Trauma. 1993; 35(3):375-82. PMID: 8371295

19.  Miller RS, Morris JA Jr, Diaz JJ Jr, Herring MB, May AK. Complications after 344 damage-control open celiotomies. J Trauma. 2005; 59(6):1365-71. PMID: 16394910 DOI: 10.1097/01.ta.0000196004.49422.af

20.  Teixeira PG, Inaba K, Dubose J, Salim A, Brown C, Rhee P, et al. Enterocutaneous fistula complicating trauma laparotomy: a major resource burden. Am Surg. 2009; 75(1):30-2. PMID: 19213393

21.  Bradley MJ, Dubose JJ, Scalea TM, Holcomb JB, Shrestha B, Okoye O, et al. Independent predictors of enteric fistula and abdominal sepsis after damage control laparotomy: results from the prospective AAST Open Abdomen registry. JAMA Surg. 2013; 148(10):947-54. PMID: 23965658 DOI: 10.1001/
jamasurg.2013.2514

22.  Pommerening MJ, DuBose JJ, Zielinski MD, Phelan HA, Scalea TM, Inaba K, et al. Time to first take-back operation predicts successful primary fascial closure in patients undergoing damage control laparotomy. Surgery. 2014; 156(2):431-8. PMID: 24962190 DOI: 10.1016/j.surg.2014.04.019

23.  Fiedler MD, Jones LM, Miller SF, Finley RK. A correlation of response time and results of abdominal gunshot wounds. Arch Surg. 1986; 121(8):902-4. PMID: 3729707 DOI: 10.1001/
archsurg.1986.01400080044007

24.  Ball CG, Williams BH, Tallah C, Salomone JP, Feliciano DV. The impact of shorter prehospital transport times on outcomes in patients with abdominal vascular injuries. J Trauma Manag Outcomes. 2013; 7(1):7-11. PMID: 24360286 DOI: 10.1186/1752-2897-7-11

25.  Ball CG. Damage control resuscitation: history, theory and technique. Can J Surg. 2014; 57(1):55-60. PMID: 24461267 DOI: 10.1503/cjs.020312

26.  CRASH-2 trial collaborators, Shakur H, Roberts I, Bautista R, Caballero J, Coats T, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet. 2010; 376(9734):
23-32. PMID: 20554319 DOI: 10.1016/S0140-6736(10)60835-5

27.  Timmermans J, Nicol A, Kairinos N, Teijink J, Prins M, Navsaria P. et al. Predicting mortality in damage control surgery for major abdominal trauma. S Afr J Surg. 2010; 48(1):6-9. PMID: 20496817

28.  Ordonez CA, Badiel M, Sanchez AI, Granados M, Garcیa AF, Ospina G, et al. Improving mortality predictions in trauma patients undergoing damage control strategies. Am Surg. 2011; 77(6):778-82. PMID: 21679650

29.  Kairinos N, Hayes PM, Nicol AJ, Kahn D. Avoiding futile damage control laparotomy. Injury. 2010; 41(1):64-8. PMID: 19570531 DOI: 10.1016/j.
injury.2009.05.036

30.  Frischknecht A, Lustenberger T, Bukur M, Turina M, Billeter A, Mica L, et al. Damage control in severely injuredtrauma patients - A ten-year experience. J Emerg Trauma Shock. 2011; 44(4):450-4. PMID: 22090736 DOI: 10.4103/0974-2700.86627

31.  Aoki N, Wall MJ, Demsar J, Zupan B, Granchi T, Schreiber MA, et al. Predictive model for survival at the conclusion of a damage control laparotomy. Am J Surg. 2000; 180(6):540-4. PMID: 11182414 DOI: 10.1016/s0002-9610(00)00497-9

32.  Brenner M, Bochicchio G, Bochicchio K, Ilahi O, Rodriguez E, Henry S, et al. Long-term impact of damage control laparotomy. Arch Surg. 2011; 146(4):395-9. PMID: 21173282 DOI: 10.1001/
archsurg.2010.284

33.  Adesanya AA, da Rocha-Afodu JT, Ekanem EE, Afolabi IR. Factors affecting mortality and morbidity in patients with abdominal gunshot wounds. Injury. 2000; 31(6):397-404. PMID: 10831735 DOI: 10.1016/s0020-1383(99)00247-8